Crankshaft sensors - can they partially fail?

Wife's 93 Chrysler 3.5L v6 had an intermittent crank sensor problem. In one sense it was a "partial" failure in that it would either cause a brief stumble or a stall, but would then work fine. But in another sense it was "all or nothing" in that the engine would COMPLETELY cut out, not just misfire or lose power. In theory, I would think the sensor could get marginal but the computer will probably always simply shut down when the signal gets too questionable.

On that car, an error code was set which allowed me to narrow it to the cam or crank sensor... I assume a Saab would have a similar error logging system but perhaps not the same ability for you to read out codes on the dashboard that the Chrysler has.

I had initially ruled the crank sensor out, wrt to a reproducible bucking problem, as my tach is rock steady - no hint of dropping - during the episodes.

But a mech suggested otherwise to me this morning, which I still have trouble believing.

I do have OBD thru the Check Engine Light (& 2 potential "No RPM" codes), but no faults

They can fail in other than an all or nothing mode. A weak signal or one that is distorted can cause a whole host of problems.

That would make it a hall effect sensor - three wires (power, ground, signal) which should output a clean square wave signal when good.

'Potential' codes? They are codes or they are not.

Do you mean 'Pending' codes? Meaning the code was recorded for a fault but not enough times consecutively to turn on the Check Engine light.

The actual codes would be helpful.

This pretty much answers my question. I hope to learn more about and buy a used o-scope in the near future.

I just meant that my Saab's built-in diagnostic has 2 "No RPM" codes, amongst about 2 dozen possible/total fault code listings.

The built-in dx system is working, but no codes are recorded.

If I may rephrase a previous statement into a another question: when a H-E sensor partially fails while driving, can a H-E sensor fed* tach remain steady?

I appreciate the help

*tach receiving input signal from ignition control unit, i.e., not a split signal: one to tach, and one from ignition control unit

This is a good idea.

It is the best way to find out what the output of the sensor actually is.

Normally I wouldn't think so, but it could.

Again, a signal may be received (a switch from high to low), but we really don't know about the quality of the signal yet. Also, the tach may have a buffer to stabilize the reading on the gauge. This could dampen out any bad pulses.

No problem.

If his vehicle had a crank sensor and a hall effect at the distributor, that would be the Cylinder IDentification sensor. It is only used to sequence the injectors and for cylinder ID during misfire detection.

I wouldn't expect it to cause a bucking problem. If it fails the system will just randomly select which injector in the firing order will go first.

(re-joining the thread after being away a few days)

I think its entirely possible. Tach needles these days seem very heavily damped, I doubt it would have any chance of responding quickly enough to show a dropped cylinder firing signal.